Archive | June, 2014

Updates from the Field – Taarifa in Iringa District

5 Jun

I am utterly exhausted as I write this, but I wanted to get info out sooner rather than later. I’ll likely come back through in a few days for cleanup.

We picked back up with Taarifa in London in, amazingly, the building it was birthed out of. Note I didn’t say conceived in. It’s a project created from deep field experience, in talking to people who live in areas with lacking water infrastructure, and seeing a gap between knowledge held in a community, the knowledge those with lots of resources to wave around have, and fulfilling the need of access to water. It’s one of the few projects I’ve seen in years of hackathons that beyond building a project to feel good, and to learn more about what response is like [which I think are both worthwhile, it’s when conflated with “field-ready” that this becomes an issue]. This one has been carried forward for years, deployed in multiple countries to various success, and always at the request of (to start AND stop) of the local government[Of course my grassroots, activist self grumbles at this]. We were back at UCL, hosted by Muki’s ExCITeS group, and things just sort of worked. More reminiscent of a hackerspace or Libre hackathon than the recent industry ones[hackathonfaq.com]. Most of the team that would be in Dar es Salaam was with us for the hackathon, and we hunkered down and sussed through code – SMS, base code, email integration, and going over again (and again) the dev environment installation documentation. This wasn’t just about getting the barebones up and going, it was also about being sure we could easily onboard new people. Because next stop – back to Dar es Salaam.

In case I haven’t said it before, I’ll say it again here – the series of hackathons were to have base infrastructure in place, and to work with local software devs to maintain their own version of the code. Doesn’t make sense that it’s in the coding language it is? Let’s change it! The last thing we want to do is design, create, and implement a “solution”[man I hate that word] from far away. Because it’s not just about the tool being used well by a community – it’s also about the community fully understanding it, and having full control over the tools they use.

The hackathon in Dar ran Saturday Sunday and involved our team of 7 people – 2 grad students for general assist, 1 core code contributor, 1 cartographer, 1 person from World Bank, 1 hardware person, and myself – and another 12 local students, software developers, and entrepreneurs. [Again, one slightly angsty point for me was that we worked closely with innovation spaces, not hack spaces. Do you know about my NOLA experience? I’ll tell you about it some time. Hilarious. But I definitely try to get as close to the root as possible when working.] We worked on the SMS implementation (yay Telerivet!), backend integration, a smartphone application[many of the water engineers have smart phones – we checked. Thrice.], and the dashboards. From the funds available through the innovation grant, we hired 3 enthusiastic and competent locals to continue working not he project over the next 3 months (at least). They’ll be joining for the biweekly Taarifa call-ins[which you can also join – will be posted to the GWOB Google Plus page once I have my wits about me – but you can see all the past ones via the same place].

Unlike many hackathons, the time limit on this one wasn’t arbitrary. We (Mark from World Bank, Austin from MoMo, George and Gregor on Taarifa, an awesome local and hackathon attendee who was able to join us named Fufiji, our amazing driver Mr Wensei, and myself – sad to see Jeremy the cartographer and Andreas, our other grad student, go away) into the car at about 19:30 and headed to our stopover in Morogoro, to continue on to Irigina the next morning. 5 hours on the road, most in traffic of a supply chain made apparent, we took up half of the dark hotel we arrived at early early in the morning. Bed nets overhead, farm animal noises out the window, we slept the sleep of exhausted travel and intensive thinking. The day after, we continued at a leisurely pace to Iringa, stopping to eat fried chicken and check out hardware stores for Austin’s MoMo. We crept along mountain roads, behind cautious trucks, their brethren crashed along the roadside and pilfered for parts. I sang Flanders and Swann, Mark sang naval songs, but we generally shared silence and admired the vastness and beauty of this country. The buildings with Xs on them are within 30 meters of the road, and are slated for demolition if cities ever grow in these rural areas – a hedged bet against the traffic nightmare of Dar. Drivers signal to each other with blinkers and high beams and horns about upcoming speed traps, safety in passing, and general hello’s. And when we finally arrived, we ate chicken and fish by the light of cell phones, the power having gone out, and talked about the next couple days.

The first day we met with district officials to explain the purpose of the innovations, so we might move forward with their blessing. It’s an incredibly hierarchical society, and so authority buy-in is essential. The district water engineer, the head of health, and of education, and even the executive director of the region come by. We demoed Taarifa, and MoMo, and everyone was much impressed. But I was left feeling.. lacking. I wanted people to understand how it worked, and how it was important for their workflow (if it was). Patience, our community partner reminded me. Tomorrow we will do activities. We go back to the hotel, and Austin solders, and I draw, and we all type and talk. The gents head out to a local bar, stopping to take part in a dancing line for a music video on the way. I stay in the hotel, knowing I will only be angry about gender politics if I venture out into night life.

The second day, I lay out a game. We have 8 water points, and the room is divided up into community members, COWSOs (Community Owned Water Source Organizations), water engineers, and the district water offices. We set some constraints – the water engineers can only visit so many water points in a round, the district should know what’s going on at the end of the round so they can properly send funds. They have data in a database about what the water points were at when last surveyed, a couple years ago. Now, GO. 10 minutes to repair anything they can. Chaos. Some points get fixed, but the water engineers went to more villages than their limit allowed (we said they were flying! Not on buses. We laugh), often to places where the points were already fixed, and all in one group. Now, reconvene and talk through it. The COWSOs were completely left out (just as they are right now! Again, laughter, but there’s something deeper there, something we don’t have time for right now), and some villages were never visited. And the district gave money to places that they had listed as broken – some of which had been fixed, and didn’t give money to some places because they thought it had been fixed but it wasn’t. We talk through distribution of money – of course places with fixed points would make good use of extra funds, but they’d prefer those with less than they have get much needed funds instead.

We redefine the parameters. Anyone can send Taarifa a message (as represented by Fufuji with a spreadsheet) about their water points. The water engineers can pay as much or as little attention to it as they want, and the district can make their report off of the information, since the platform will have seen the updates to the water. We go through the exercise again, and this time there is clarity. We talk through it. Resources were better used. Time was better used. Everyone felt heard[It was, in all honesty, pretty awesome]. Then we take tea, demo the actual tools, and head out to the field.

Austin fit an exposed pipe with a MoMo, updating the firmware from the field. First time I’ve seen someone code from a pit, hands dirty, umbrella held overhead to shield the screen. We struggle with what SIM card to use in this spot, how to test given water hasn’t flown since morning, and that a 1” pipe was promised but a 3/4” pipe is what we’ve got. It finally works, tape and adapters and dirt-laden code, everyone sweaty but high on possibility. If it will send updates under the 2.5 feet of dirt which will also protect the device and the pipe from elements and scrappers, is another question. We send some messages via Taarifa, and thank everyone, and acknowledge George, Gregor, and Austin will be back within a week. Reiterate that it is the combination of the device and the people that ensure a robust system.

And then we have to get in the car and race the sun and the mountains and the supply chain on dangerous curves so I might catch my flight. An 11 hour care ride later, I’m sitting in an airport, queuing up my laptop to write this, exhausted but happy.

Dar es Salaam Hackathon updates

1 Jun

Start at 0:0:50

Day 1 Report Out

SMS integration with Telerivet. It works! Translated into Swahili. Down to 10 lines of code. If you send something in the wrong format, it requests it in the right format, with example. Next up is #16 – goes from API to do something.

Translation – if we build this out, it should be usable to the community.

Discussed the user interface, have a few ideas on how to make it as simple as possible and usable for people who don’t know about coding or what’s behind it. Start coding tomorrow!

Setting up a new API end point. Spent most of the day understanding how it works, and how to make it work the way I want it to.

Community partners talked about how to indicate where a water point is, outside of its super long ID number, which is usually missing. USSD interface possibilities.

Data for flow rates are going to the cloud, working on how it ties into Taarifa in the model, and where it should live.

Built a mobile app. Have a fronted with the Tanzanian flag! Will code it up tomorrow. Can also add in new water points.

Day 2 Gear Up and Data Flow

The internet is really bad here, so here’s a link to the video, as I can’t watch it.